Valence Band Control of Metal Silicide Films via Stoichiometry - The

Jun 20, 2016 - (24) The intensity and shape of XPS VB spectra of open d shell metals, such as Pt, are modified and therefore unable to resemble DOS ca...
0 downloads 6 Views 2MB Size
Letter pubs.acs.org/JPCL

Valence Band Control of Metal Silicide Films via Stoichiometry Frank Streller,† Yubo Qi,‡ Jing Yang,‡ Filippo Mangolini,†,§ Andrew M. Rappe,†,‡ and Robert W. Carpick*,†,¶ †

Department of Materials Science and Engineering, ‡Makineni Theoretical Laboratories, Department of Chemistry, and ¶Department of Mechanical Engineering and Applied Mechanics, University of Pennsylvania, Philadelphia, Pennsylvania 19104, United States ABSTRACT: The unique electronic and mechanical properties of metal silicide films render them interesting for advanced materials in plasmonic devices, batteries, fieldemitters, thermoelectric devices, transistors, and nanoelectromechanical switches. However, enabling their use requires precisely controlling their electronic structure. Using platinum silicide (PtxSi) as a model silicide, we demonstrate that the electronic structure of PtxSi thin films (1 ≤ x ≤ 3) can be tuned between metallic and semimetallic by changing the stoichiometry. Increasing the silicon content in PtxSi decreases the carrier density according to valence band X-ray photoelectron spectroscopy and theoretical density of states (DOS) calculations. Among all PtxSi phases, Pt3Si offers the highest DOS due to the modest shift of the Pt5d manifold away from the Fermi edge by only 0.5 eV compared to Pt, rendering it promising for applications. These results, demonstrating tunability of the electronic structure of thin metal silicide films, suggest that metal silicides can be designed to achieve application-specific electronic properties.

etal silicide (MexSi) thin films were extensively studied in the 1980s and have found use in the microelectronics industry as materials for electronic contacts, local interconnects, and diffusion barriers.1 Recently, MexSi regained scientific attention as they are considered candidate materials for a variety of novel applications, such as plasmonics,2 lithium-ion batteries,3 field emitters,4−6 thermoelectrics,7−10 field-effect transistors,11 and nanoelectromechanical switches.12−17 The increasing popularity of MexSi is a consequence of their metallike electrical properties, semiconductor-like thermal transport, mechanical robustness, and thermal stability. However, Boltasseva et al.2 and Cheng et al.,3 among others, pointed out that the composition and properties of MexSi need to be carefully optimized to fully utilize their potential in nextgeneration applications. We recently presented a methodology to precisely tune the composition of MexSi thin films by means of controlled solid-state diffusion.15 This allows for the formation of platinum silicide (PtxSi) thin films over a wide composition range (1 ≤ x ≤ 3), including the novel Pt3Si stoichiometry, with an associated wide range of mechanical and electronic properties.12,14,15 While it is known that the MexSi composition strongly affects the resulting electrical, mechanical, and adhesive properties,12,14 systematic studies determining the effect of MexSi composition on specific properties are lacking. Here, we compare valence band (VB) X-ray photoelectron spectroscopy (XPS) measurements with calculations of the density of states (DOS) around the Fermi edge to elucidate the relationship between composition and electronic structure/ properties for PtxSi thin films. Using source-limited solid-state diffusion (i.e., films are formed by annealing of sequentially deposited Pt and a-Si thin films of specific thickness ratios),15 we fabricated PtxSi thin films of Pt3Si, Pt2Si, and PtSi. These films were interrogated

M

© XXXX American Chemical Society

using high-resolution VB XPS measurements to determine the DOS at the Fermi edge, which accurately determines the electronic structure and many of the electronic properties of these PtxSi films. A high DOS at the Fermi edge is a requirement for having high electrical conductivity.11 The Fermi edge shape therefore provides an indication of the electronic character of the tested material (metal, semimetal, or semiconductor), as shown schematically in Figure 1a. Metals are characterized by a large quantity of charge carriers (high DOS) near the Fermi edge that allows them to be efficient electrical conductors. Semimetals possess a significantly reduced DOS at the Fermi edge without an energy band gap (which, combined with their typically higher Seebeck coefficient and lower thermal conductivities compared to metals, makes them suitable thermoelectric materials).18 The Fermi edges of semiconductors are located inside of the band gap without any electronic states at the edge. Because XPS probes the occupied states of a material (see Figure 1b), the spectral features of a VB XPS spectrum directly correlate with the DOS. This relationship allows us to directly compare XPS measurements with theoretical DOS calculations. Figure 2 shows the results of the VB XPS measurements and the theoretical DOS calculations for [Pt], [Pt3Si], [Pt2Si], and [PtSi] films (the “[...]” notation refers to the dominant phase of four respective mixtures achieved experimentally, as shown in Figure 2c). The composition of the produced PtxSi films was determined using quantitative XPS (see the Experimental and Computational Details section). This demonstrated the phase selectivity as shown in Figure 2c; the [Pt3Si] film was 74% Received: April 14, 2016 Accepted: June 20, 2016

2573

DOI: 10.1021/acs.jpclett.6b00799 J. Phys. Chem. Lett. 2016, 7, 2573−2578

Letter

The Journal of Physical Chemistry Letters

Figure 1. Fermi edge shapes and locations in XPS spectra. (a) Schematic Fermi edge shape of a metal, a semimetal, and a semiconductor. The intensity of charge carriers at the Fermi edge reduces from metals to semimetals and becomes zero for semiconductors. (b) Location of the Fermi edge within an experimental Pt XPS spectrum. The Fermi edge is located at a binding energy value of 0 eV.

Figure 2. Experimental and theoretical Fermi edge shapes and compositions for [Pt], [Pt3Si], [Pt2Si], and [PtSi] samples. (a) Experimental Fermi edge shapes of [Pt], [Pt3Si], [Pt2Si], and [PtSi] samples. The VB electron population decreases with increasing Si content of the silicide. (b) Theoretical Fermi edge shapes of [Pt], [Pt3Si], [Pt2Si], and [PtSi] samples. The VB electron population decreases with increasing Si content of the silicide. The dotted vertical lines represent the location of the Pt5d manifold for PtSi (2.7 eV below the Fermi edge), Pt2Si (1.8 eV below the Fermi edge), and Pt3Si (0.5 eV below the Fermi edge). (c) Thin film composition of [Pt], [Pt3Si], [Pt2Si], and [PtSi] samples as determined from quantitative XPS measurements.

Pt3Si, the [Pt2Si] film was 70% Pt2Si, the [PtSi] film was 92% PtSi, and the [Pt] film was 100% Pt. These exact compositions were also employed in the theoretical DOS calculations using a linear combination approach to ensure a better comparability of the results. The VB XPS measurements show a direct correlation between the Si concentration and the Fermi edge shape (Figure 2a). This agrees with recent work by Fryer and Lad, who performed VB XPS measurements of co-deposited PtxSi films (1 ≤ x ≤ 3).19 The experimental DOS at the Fermi edge systematically decreases with increasing Si concentration from Pt toward PtSi. The analyzed films show metallic and reduced-metallic Fermi edge shapes in the case of Pt and Pt3Si, respectively, and typical semimetallic behavior for the Pt2Si and PtSi films. The theoretical DOS calculations show a similar

reduction of the DOS at the Fermi edge from Pt to PtSi. Additionally, the calculated DOS for Pt, Pt2Si, and PtSi are similar to those reported in previous works by Bentmann et al.20 and Franco et al.21 No DOS calculations for Pt3Si have been previously reported in the literature. Figure 3 shows a direct comparison of the experimental and theoretical DOS for [Pt], [Pt3Si], [Pt2Si], and [PtSi] films. Overall, we find that all experimental PtxSi VB spectra capture not only the general shape but also several individual features of the theoretical PtxSi DOS very well (Figure 3b−d). However, Figure 3a shows that the experimental Pt VB spectrum is unable to resemble the structural detail of the theoretical Pt DOS. Comparisons of our XPS with other XPS data show excellent agreement;19,22−24 similarly, comparisons of our 2574

DOI: 10.1021/acs.jpclett.6b00799 J. Phys. Chem. Lett. 2016, 7, 2573−2578

Letter

The Journal of Physical Chemistry Letters

Figure 3. Comparison between experimental and theoretical Fermi edge shapes for [Pt], [Pt3Si], [Pt2Si], and [PtSi] samples. (a) [Pt] sample (100% metallic Pt). (b) [Pt3Si] sample (which includes 74% Pt3Si). (c) [Pt2Si] sample (which includes 70% Pt2Si). (d) [PtSi] sample (which includes 92% PtSi).

Pt5d manifold of only approximately 0.5 eV, which results in a metallic-like character of the Fermi edge. Table 1 quantitatively compares the experimental and calculated DOS of Pt, Pt3Si, Pt2Si, and PtSi by means of their

theoretical DOS and other DOS calculations show excellent agreement,24,25 but there is some disagreement between the two for the case of Pt. This points toward an inherent difficulty in matching the VB XPS spectrum of Pt to its theoretical DOS counterpart. The reasons for this phenomenon are discussed by Goldmann et al.22 and Hofmann et al.24 The intensity and shape of XPS VB spectra of open d shell metals, such as Pt, are modified and therefore unable to resemble DOS calculations due to (1) instrumental resolution of the XPS system, (2) matrix element modulation across the width of the d band, (3) the lifetime of the photohole, (4) interaction of the photohole with the conduction electrons, and (5) ineleastic electron scattering. The factor that most prominently influences the structural detail of the VB spectra is the lifetime of the photohole, which leads to a broadening of the XPS peaks. These effects are less pronounced in PtxSi due to the filling up of the d shell due to silicidation, in agreement with our results. The measured and calculated VBs for Pt and the PtxSi films are dominated by the Pt5d manifold. The Pt5d position within the VB greatly influences the observed Fermi edge shapes and consequently the carrier densities. A maximum of the Pt5d manifold is located directly at the Fermi edge in the case of Pt (Figure 3a), whereas it shifts further away from the Fermi edge for PtxSi with increasing Si content. Our calculations show that the Pt5d manifold of PtSi is shifted to approximately 2.7 eV below the Fermi edge, whereas the Pt5d manifold of Pt2Si is only shifted to approximately 1.8 eV below the Fermi edge (see the dotted vertical lines in Figure 2b), similar to previously reported values.20,21 The novel Pt3Si films show a shift of the

Table 1. Normalized Carrier Densities Computed from VB XPS Experiments and DOS Calculations and Measured Sheet Resistance Values normalized carrier density Pt Pt3Si Pt2Si PtSi

VB XPS

DOS calculation

sheet resistance (Ω/□)

1 0.59 0.20 0.17

1 0.61 0.28 0.19

2.6 18.9 31.8 57.6

normalized carrier densities. The carrier densities were obtained by integrating the experimental and theoretical DOS within a 2kBT (∼0.05 eV for T = 300 K) energy window around the Fermi edge.11 The normalized carrier densities computed from the VB XPS measurements are in excellent agreement with the DOS calculations and confirm the systematic decrease in carrier density with increasing Si content. The PtSi and Pt2Si films were found to possess approximately 17−19% and 20−28% of the carrier density of Pt, respectively. These values are in good agreement with theoretical calculations performed by Bentmann et al.20 The Pt-rich Pt3Si film possesses a very high carrier density of 59−61% of the DOS of Pt, which is a more than 3fold improvement over the PtSi carrier density. 2575

DOI: 10.1021/acs.jpclett.6b00799 J. Phys. Chem. Lett. 2016, 7, 2573−2578

Letter

The Journal of Physical Chemistry Letters

properties of other metal silicides based on Ni, Cu, Au, Fe, Co, Pd, Ag, and Rh among others. We expect that the open d shell metals (e.g., Fe, Co, Pd) will exhibit the strongest tunability of the electrical properties due to their similarity with Pt, which was shown to have a strong compositional dependence on the Pt5d manifold position.

As mentioned above, a high DOS at the Fermi level is one of the requirements for having a high electrical conductivity. We performed four point probe measurements on the Pt, Pt3Si, Pt2Si, and PtSi films to determine the sheet resistance as an inverse measure of the electrical conductivity (Figure 4). The



EXPERIMENTAL AND COMPUTATIONAL DETAILS Thin Film Formation. To form the PtxSi films, Pt and a-Si films were sputter-deposited in a Denton Vacuum Explorer 14 sputterer (Denton Vacuum Inc., Moorestown, NJ) with a purity of 99.99% for both films and subsequently annealed under high vacuum (10−7 Torr) at 600 °C for a hold time of 10 min (reached after using a heating rate of 30 °C/min). Pt was deposited in dc mode at 450 W and a-Si in ac mode at 230 W. The thicknesses of the Pt and a-Si films were chosen to obtain nearly pure phases of Pt3Si, Pt2Si, and PtSi upon annealing.15 The Pt and a-Si depositions were conducted sequentially in the same deposition system under maintained vacuum. This minimizes contaminant adsorption between the layers and oxidation of the a-Si, both of which are inhibiting factors for silicidation. The PtxSi film thicknesses were between 40 and 50 nm for all films studied here. XPS Analysis. The chemistry of the near-surface region was investigated by XPS using a customized XPS spectrometer (VG Scienta AB, Uppsala, Sweden).28 XPS analyses were performed using a monochromatic Al Kα source (photon energy of 1486.6 eV). The residual pressure in the analysis chamber was consistently less than 1 × 10−8 Torr. The spectrometer was calibrated according to ISO 15472:2001 with an accuracy of ±0.05 eV. Survey and high-resolution spectra were acquired in constant analyzer energy mode with pass energies of 200 and 100 eV, respectively. The full width at half-maximum (fwhm) of the peak height for the high-resolution Ag 3d5/2 signal of a sputter-cleaned Ag sample was 0.57 eV. The spectra were processed using CasaXPS software (v.2.3.16, Casa Software Ltd., Wilmslow, Cheshire, U.K.). Background subtraction was performed using the Shirley−Sherwood method. The quantitative evaluation of XPS data, as described in ref 29, was based on integrated intensity using a first-principles model and applying Powell’s equation. The inelastic mean free path was calculated using the TPP-2M formula.30 Curve synthesis for the Pt 4f peaks was performed by constraining the integrated intensity ratio of these two signals to 3:4 and their energy separation to 3.33 eV. The reference energies for Pt 4f7/2 peaks were 71.05, 71.55, 72.18, and 72.75 eV for Pt, Pt3Si, Pt2Si, and PtSi, respectively, and are in agreement with literature values.12,15 The Pt 4f peaks have been chosen to determine sample surface chemistry due to their high intensity and the high sensitivity of their position to the PtxSi stoichiometry, in contrast with the less intense and less stoichiometrically sensitive position of the Si 2p peaks. Theoretical DOS Calculations. DFT calculations were performed on PtSi, Pt2Si, Pt3Si, and Pt with generalized gradient approximation31 (GGA) exchange−correlation functionals implemented in QUANTUM ESPRESSO package.32 We used norm-conserving plane-wave pseudopotentials for all of the species.33,34 The kinetic energy cutoff to the wave function expansion was Ecut= 680 eV. An 8 × 8 × 8 Monkhorst−Pack35 grid k−point mesh was used to sample the Brillouin zone (BZ) in structural optimization, while a denser one, 24 × 24 × 24, was used for DOS calculation. The

Figure 4. Sheet resistance of [Pt], [Pt3Si], [Pt2Si], and [PtSi] samples.

measured sheet resistance values show a systematic decrease in electrical conductivity from Pt to PtSi, which confirms the expected qualitative trend of the carrier density. The sheet resistance values for Pt, Pt3Si, Pt2Si, and PtSi were measured to be 2.6, 18.9, 31.8, and 57.6 Ω/□, respectively. The corresponding resistivity values were calculated to be 10.6, 75.5, 127.2, and 230.4 μΩ−cm, respectively. The obtained sheet resistance and resistivity values for Pt, Pt2Si, and PtSi are in good agreement with results reported in other works.26,27 The high electrical conductivity (and, correspondingly, the low sheet resistance and resistivity) of Pt3Si compared to that of Pt2Si and PtSi could motivate the use of Pt3Si in several applications as a replacement for Pt2Si or PtSi. Most notably for the semiconductor industry, where PtSi is considered as an attractive contact material to the source, drain, and gate for CMOS field effect transistors because of its low Schottky barrier and high thermal stability. However, PtSi suffers from low electrical conductivity due to its low DOS, which this work has verified. Recent work by Slepko and Demkov investigates Ti doping of PtSi in an attempt to increase its DOS and thereby its electrical conductivity.11 The researchers succeeded in increasing the DOS of PtSi by approximately 1.7 times through alloying with 12.5 atom % Ti. However, while the Ti doping increased the DOS, it also introduced Ti impurities that act as scattering centers and decrease the electrical conductivity. Here we were able to show that the Pt3Si phase could solve these issues by inherently possessing a high DOS near the Fermi edge (approximately 3.4 times higher than the DOS of PtSi) and a low sheet resistance and resistivity, thus indicating a high electrical conductivity. Because no doping is necessary to achieve this high DOS, Pt3Si does not suffer from additional creation of scattering centers. In summary, we show that the electronic structure and properties of PtxSi thin films can be tuned for specific applications between metallic and semimetallic properties by controlling the film stoichiometry. The comparison of VB XPS spectra with theoretical density functional theory (DFT) calculations shows that the DOS near the Fermi edge of Pt3Si is significantly higher than that of Pt2Si and PtSi. The resulting high electrical conductivity of Pt3Si makes this stoichiometry particularly interesting for applications that demand high electrical conductivity combined with high thermal and mechanical stability.12 This work on PtxSi provides a framework for studying the tunability of the electrical 2576

DOI: 10.1021/acs.jpclett.6b00799 J. Phys. Chem. Lett. 2016, 7, 2573−2578

Letter

The Journal of Physical Chemistry Letters

(7) Rowe, D. M. Thermoelectrics Handbook: Macro to Nano; Taylor & Francis Group: Abingdon, U.K., 2006. (8) Fedorov, M. I. Thermoelectric silicides: Past, present and future. J. Thermoelectr. 2009, 2, 51−60. (9) Itoh, T.; Yamada, M. Synthesis of thermoelectric manganese silicide by mechanical alloying and pulse discharge sintering. J. Electron. Mater. 2009, 38, 925−929. (10) de Boor, J.; Dasgupta, T.; Kolb, H.; Compere, C.; Kelm, K.; Mueller, E. Microstructural effects on thermoelectric efficiency: A case study on magnesium silicide. Acta Mater. 2014, 77, 68−75. (11) Slepko, A.; Demkov, A. A. Band engineering in silicide alloys. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 035311. (12) Streller, F.; Wabiszewski, G. W.; Mangolini, F.; Feng, G.; Carpick, R. W. Tunable, source-controlled formation of platinum silicides and nanogaps from thin precursor films. Adv. Mater. Interfaces 2014, 1, 1300120. (13) Loh, O. Y.; Espinosa, H. D. Nanoelectromechanical contact switches. Nat. Nanotechnol. 2012, 7, 283−295. (14) Streller, F.; Wabiszewski, G. E.; Carpick, R. W. Next-generation nanoelectromechanical switch contact materials. IEEE Nanotechnol. Mag. 2015, 9, 18−24. (15) Streller, F.; Agarwal, R.; Mangolini, F.; Carpick, R. W. Novel metal silicide thin films by design via controlled solid-state diffusion. Chem. Mater. 2015, 27, 4247−4253. (16) Streller, F.; Wabiszewski, G. E.; Durham, D. B.; Yang, F.; Yang, J.; Qi, Y.; Srolovitz, D. J.; Rappe, A. M.; Carpick, R. W. Novel materials solutions and simulations for nanoelectromechanical switches. Proceedings of the IEEE 61st Holm Conference on Electrical Contacts; San Diego, CA, 2015; pp 363−369. Doi: 10.1109/ HOLM.2015.7355122. (17) Streller, F.; Wabiszewski, G. E.; Carpick, R. W. Development and assessment of next-generation nanoelectromechanical switch contact materials. Proceedings of the 14th International Conference on Nanotechnology; Toronto, ON, 2014; pp 141−145. Doi: 10.1109/ NANO.2014.6967966. (18) Bubnova, O.; Khan, Z. U.; Wang, H.; Braun, S.; Evans, D. R.; Fabretto, F.; Hojati-Talemi, P.; Dagnelund, D.; Arlin, J.-P.; Geerts, Y. H.; et al. Semi-metallic polymers. Nat. Mater. 2013, 13, 190−194. (19) Fryer, R. T.; Lad, R. J. Synthesis and thermal stability of Pt3Si, Pt2Si, and PtSi films grown by e-beam co-evaporation. J. Alloys Compd. 2016, 682, 216−224. (20) Bentmann, H.; Demkov, A. A.; Gregory, R.; Zollner, S. Electronic, optical, and surface properties of PtSi thin films. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 78, 205302. (21) Franco, N.; Klepeis, J. E.; Bostedt, C.; Van Buuren, T.; Heske, C.; Pankratov, O.; Terminello, L. J. Valence band study of PtSi by synchrotron radiation photoelectron spectroscopy. J. Electron Spectrosc. Relat. Phenom. 2001, 114−116, 1191−1196. (22) Hoechst, H.; Huefner, S.; Goldmann, A. XPS-valence bands of iron, cobalt, palladium and platinum. Phys. Lett. A 1976, 57, 265−266. (23) Strasser, P.; Koh, S.; Anniyev, T.; Greeley, J.; More, K.; Yu, C.; Liu, Z.; Kaya, S.; Nordlund, D.; Ogasawara, H.; Toney, M. F.; Nilsson, A. Lattice-strain control of the activity in dealloyed core-shell fuel cell catalysts. Nat. Chem. 2010, 2, 454−460. (24) Hofmann, T.; Yu, T. H.; Folse, M.; Weinhardt, L.; Baer, M.; Zhang, Y.; Merinov, B. V.; Myers, D. J.; Goddard, W. A., III; Heske, C. Using photoelectron spectroscopy and quantum mechanics to determine d-band energies of metals for catalytic applications. J. Phys. Chem. C 2012, 116, 24016−24026. (25) Krupski, K.; Moors, M.; Jozwik, P.; Kobiela, T.; Krupski, A. Structure determination of Au on Pt(111) surface: LEED, STM and DFT study. Materials 2015, 8, 2935−2952. (26) Powell, R. W.; Tye, R. P.; Woodman, M. J. Thermal conductivities and electrical resistivities of the platinum metals. Platinum Met. Rev. 1962, 6, 138−143. (27) Conforto, E. Formation and properties of nanometer-thick platinum silicide layers. Ph.D. Thesis, Ecole Polytechnique Federale Du Lausanne, 1996.

occupation of the states around the Fermi energy EF was calculated from the Fermi−Dirac distribution36 f (E ) =

1 (E − E F)/ kT

e

+1

where f(E) is the occupation probability for the state with energy E, k is the Boltzmann constant, and T is the temperature, which was selected to be 298 K. The exact experimental compositions, that is, as reported in Figure 2c, were employed in the DOS calculation to ensure better comparability. The DOS for a nonpure compound was estimated from a linear combination of the DOS of each of its components. Examples of DFT DOS calculations using a similar methodology can be found in refs 37−40.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 205-898-4608. Present Address §

F.M.: Institute of Functional Surfaces, School of Mechanical Engineering, University of Leeds, LS2 9JT Leeds, United Kingdom. Author Contributions

F.S. prepared the samples and carried out the XPS experiments and resistivity measurements. Y.Q. and J.Y. performed the DOS calculations. A.M.R. oversaw the DOS calculations. F.M oversaw the XPS measurements. R.W.C. supervised the research. F.S. wrote the manuscript. Y.Q., J.Y., F.M., A.M.R., and R.W.C. edited the manuscript. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors acknowledge the use of instrumentation from the Nano/Bio Interface Center (NBIC) at the University of Pennsylvania. Funding from the DMREF program of the National Science Foundation (DMREF-1334241) is acknowledged. F.M. acknowledges support from the Marie Curie International Outgoing Fellowship for Career Development within the 7th European Community Framework Program under Contract No. PIOF-GA-2012-328776. The authors thank R. Agarwal and P. Nukala for instrumental support. A.M.R. acknowledges support from the Department of Energy Office of Basic Energy Sciences, under grant number DE-FG0207ER15920. The authors thank the National Energy Research Scientific Computing Center of the Department of Energy.



REFERENCES

(1) Miglio, L.; d’Heurle, F. Silicides - Fundamentals and Applications; World Scientific Publishing: Singapore, 2000. (2) Naik, G. V.; Shalaev, V. M.; Boltasseva, A. Alternative plasmonic materials: Beyond gold and silver. Adv. Mater. 2013, 25, 3264−3294. (3) Liu, C.; Li, F.; Ma, L.-P.; Cheng, H.-M. Advanced materials for energy storage. Adv. Mater. 2010, 22, E28−E62. (4) Lin, H.-K.; Cheng, H.-A; Lee, C.-Y.; Chiu, H.-T. Chemical vapor deposition of TiSi nanowires on C54 TiSi2 thin film: An amorphous titanium silicide interlayer assisted nanowire growth. Chem. Mater. 2009, 21, 5388−5396. (5) Li, H.; Wu, J.; Wang, Z. M. Silicon-Based Nanomaterials; Springer: Berlin, 2013. (6) Higgins, J. M.; Ding, R.; Jin, S. X. Synthesis and characterization of manganese-rich silicide (α-Mn5Si3, β-Mn5Si3, and β-Mn3Si) nanowires. Chem. Mater. 2011, 23, 3848−3853. 2577

DOI: 10.1021/acs.jpclett.6b00799 J. Phys. Chem. Lett. 2016, 7, 2573−2578

Letter

The Journal of Physical Chemistry Letters (28) Mangolini, F.; Ahlund, J.; Wabiszewski, G. E.; Adiga, V. P.; Egberts, P.; Streller, F.; Backlund, K.; Karlsson, P. G.; Wannberg, B.; Carpick, R. W. Angle-resolved environmental X-ray photoelectron spectroscopy: A new laboratory setup for photoemission studies at pressures up to 0.4 Torr. Rev. Sci. Instrum. 2012, 83, 093112. (29) Mangolini, F.; Rossi, A.; Spencer, N. D. Chemical reactivity of triphenyl phosphorothionate (TPPT) with iron: An ATR/FT-IR and XPS investigation. J. Phys. Chem. C 2011, 115, 1339−1354. (30) Tanuma, S. Surface Analysis by Auger and X-ray Photoelectron Spectroscopy; IM Publications: Charlton, U.K., 2003. (31) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (32) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; et al. Quantum ESPRESSO: A modular and open−source software project for quantum simulations of materials. J. Phys.: Condens. Matter 2009, 21, 395502−395520. (33) Ramer, N. J.; Rappe, A. M. Designed nonlocal pseudopotentials for enhanced transferability. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 59, 12471−12478. (34) Rappe, A. M.; Rabe, K. M.; Kaxiras, E.; Joannopoulos, J. D. Optimized pseudopotentials. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41 (2), 1227. (35) Monkhorst, H. J.; Pack, J. D. Special points for Brillouin−zone integrations. Phys. Rev. B 1976, 13, 5188−5192. (36) Kittel, C. Introduction to Solid State Physics; John Wiley & Sons, Inc.: Toronto, Canada, 1971. (37) Brehm, J. A.; Takenaka, H.; Lee, C.-W.; Grinberg, I.; Bennett, J. W.; Rutenberg Schoenberg, M.; Rappe, A. M. Density functional theory study of hypothetical PbTiO3-based oxysulfides. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 89, 195202. (38) Martirez, J. M. P.; Morales, E. H.; Saidi, W. A.; Bonnell, D. A.; Rappe, A. M. Atomic and electronic structure of the BaTiO3(001) (√5 x√5) R26.6°. Phys. Rev. Lett. 2012, 109, 256802. (39) Gou, G. Y.; Bennett, J. W.; Takenaka, H.; Rappe, A. M. Post density functional theoretical studies of highly polar semiconductive Pb(Ti1−xNix)O3−x solid solutions: Effects of cation arrangement on band gap. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 83, 205115. (40) Grinberg, I.; Suchomel, M. R.; Dmowski, W.; Mason, S. E.; Wu, H.; Davies, P. K.; Rappe, A. W. Structure and polarization in the high Tc ferroelectric Bi(Zn, Ti)O3-PbTiO3 solid solutions. Phys. Rev. Lett. 2007, 98, 107601.

2578

DOI: 10.1021/acs.jpclett.6b00799 J. Phys. Chem. Lett. 2016, 7, 2573−2578